pymc.adagrad_window(loss_or_grads=None, params=None, learning_rate=0.001, epsilon=0.1, n_win=10)[source]#

Returns a function that returns parameter updates. Instead of accumulated estimate, uses running window

loss_or_grads: symbolic expression or list of expressions

A scalar loss expression, or a list of gradient expressions

params: list of shared variables

The variables to generate update expressions for

learning_rate: float

Learning rate.

epsilon: float

Offset to avoid zero-division in the normalizer of adagrad.

n_win: int

Number of past steps to calculate scales of parameter gradients.


A dictionary mapping each parameter to its update expression